basically all I need to know is what are the standard methods to achieve the below.
So, I have a fuzzy set A containing (say) four elements. For each element I have a degree of membership. The degrees of membership sum to one. I want to get a crisp version of A. I'm aware we have alpha-cuts. But is there any standard method to find alpha?
(in a nutshell: the elements are actually features in a data set, and the degrees of membership is how relevant each feature is. I'm trying to do feature selection here.)
Thanks!